Getting Things Done on Computational RFIDs with Energy-Aware - PowerPoint PPT Presentation

HotPower ’08 Getting Things Done ☑ on Computational RFIDs with Energy-Aware Checkpointing and Voltage-Aware Scheduling Benjamin Ransford , Shane Clark, Mastooreh Salajegheh, Kevin Fu Department of Computer Science University of Massachusetts Amherst

Scenario: RFID Sensor Network [HotNets ’08] • Maintenance-free • Batteryless nodes • RF power harvesting • Try to do public-key crypto. Photos: Impinj, M. W. Moss Ltd., HotPower ’08 — Ben Ransford — 2 reinforcedearth.com

Scenario: RFID Sensor Network [HotNets ’08] • Maintenance-free • Batteryless nodes • RF power harvesting • Try to do public-key crypto. Photos: Impinj, M. W. Moss Ltd., HotPower ’08 — Ben Ransford — 2 reinforcedearth.com

Scenario: RFID Sensor Network [HotNets ’08] • Maintenance-free • Batteryless nodes • RF power harvesting • Try to do public-key crypto. Photos: Impinj, M. W. Moss Ltd., HotPower ’08 — Ben Ransford — 2 reinforcedearth.com

The next 15 minutes 1. Batteryless computing with computational RFID (CRFID) 2. Obstacles to computing on harvested energy • Fluctuating supply, power loss 3. Mementos: s/w for getting things done • Checkpointing, program reordering HotPower ’08 — Ben Ransford — 3

Batteries constrain design. Big & heavy relative to circuits. Must be replaced or recharged. Energy density slooooowly increasing. (1991: 204 Wh/l ... 2005: 514 Wh/l) photos: Duracell, Micro-Tools.com, Crossbow HotPower ’08 — Ben Ransford — 4

How can we do useful computation without a battery? HotPower ’08 — Ben Ransford — 5

How can we do useful computation without a battery? Focus on energy harvesting. HotPower ’08 — Ben Ransford — 5

Perils of RF harvesting • Devices become dependent on energy supply • Unpredictable supply • Fluctuating voltage • Frequent loss of power/state Photo: Lois Elling HotPower ’08 — Ben Ransford — 6

Today’s batteryless computers RFID tag smart card non-programmable must finish in one circuitry energy lifecycle Photos: thisismoney.co.uk, TI HotPower ’08 — Ben Ransford — 7

Computational RFID (new term) • Modern ultra-low-power (1.5µA sleep, 600µA active) programmable microcontroller • von Neumann architecture • RAM, flash memory e.g. No battery... WISP RF harvesting. [IEEE TIM ’08] HotPower ’08 — Ben Ransford — 8

Computational RFID (new term) • Modern ultra-low-power (1.5µA sleep, 600µA active) programmable microcontroller • von Neumann architecture • RAM, flash memory e.g. No battery... WISP RF harvesting. [IEEE TIM ’08] HotPower ’08 — Ben Ransford — 8

RFID: Power + Query Reader 0x3C6B23A4 Static ID Computational RFID: HotPower ’08 — Ben Ransford —

RFID: Power + Query Reader 0x3C6B23A4 Static ID Computational RFID: Power + Query Reader 0x1234CAFE Results of computation or sensing HotPower ’08 — Ben Ransford —

Perils of RF harvesting • Devices become dependent on energy supply • Unpredictable supply • Fluctuating voltage • Frequent loss of power/state Photo: Lois Elling HotPower ’08 — Ben Ransford — 10

Perils of RF harvesting • Devices become dependent on energy supply • Unpredictable supply • Fluctuating voltage We can • Frequent loss of power/state address these. Photo: Lois Elling HotPower ’08 — Ben Ransford — 10

Getting things done Major goal: help programs on CRFIDs make forward progress despite fluctuating voltage and constant interruption. HotPower ’08 — Ben Ransford — 11

Our system: Mementos • Designed to aid forward progress. • Execution checkpointing (suspend, resume) • Program reordering HotPower ’08 — Ben Ransford — 12

Our system: Mementos • Designed to aid forward progress. • Frequent loss of power/state • Execution checkpointing (suspend, resume) • Program reordering HotPower ’08 — Ben Ransford — 12

Our system: Mementos • Designed to aid forward progress. • Frequent loss of power/state • Execution checkpointing (suspend, resume) • Program reordering • Fluctuating voltage HotPower ’08 — Ben Ransford — 12

Checkpointing • Frequent loss of power/state • Idea: save state to flash before dying • Problem: flash writes consume significant energy when it’s least available. • Flash vs. register: 400x more energy • Flash vs. memory: 40x more energy HotPower ’08 — Ben Ransford — 13

Checkpointing • Compile time static analysis: • Compute per-block energy estimates • Run time: • CRFID checks own voltage • Dynamic checkpointing decision HotPower ’08 — Ben Ransford — 14

Energy estimation at compile time label1: MOV R11, R12 1 nJ ADD R12, R8 1 nJ (Flash write) 461 nJ JMP label2 -- ... ... Platform-specific Annotated energy profile instruction stream HotPower ’08 — Ben Ransford — 15

e.g.: modexp • Halve 32-bit exponent, square 32-bit base • No checkpointing: dies before finishing HotPower ’08 — Ben Ransford — 16

e.g.: modexp • Halve 32-bit exponent, square 32-bit base • No checkpointing: dies before finishing HotPower ’08 — Ben Ransford — 16

e.g.: modexp • Halve 32-bit exponent, square 32-bit base • No checkpointing: dies before finishing • Checkpoint halfway through: • Save base, exp., accumulated result after 15 iterations; die before finishing • Restore from checkpoint; 17 more iterations; complete. HotPower ’08 — Ben Ransford — 16

Program reordering • Fluctuating voltage • Observations: • Some operations require higher voltage • Voltage tends to decline during each device lifecycle • Microcontrollers don’t like continuously varying voltage (PLL logic limitations) HotPower ’08 — Ben Ransford — 17

Program reordering • Static analysis at compile time • Estimate energy requirements • Derive dependency graph • Must not violate program semantics! HotPower ’08 — Ben Ransford — 18

Program reordering • Voltage declines: reorder independent code chunks at compile time to execute high-V ops when voltage is high I4 I4 Flash Min. Required Voltage Min. Required Voltage write block I2 I5 I5 I2 I3 I3 I1 I1 Before After HotPower ’08 — Ben Ransford — 19

Program reordering • Smaller timescale: adaptively reschedule program chunks at run time to avoid logjams Voltage These never get to run! { Run queue HotPower ’08 — Ben Ransford — 20

Challenges • Predicting program behavior is hard. • Balance checkpointing behavior: • How much state to save • How often to checkpoint • Program reordering: • Finding dependencies can be hard HotPower ’08 — Ben Ransford — 21

Physical barriers • Can’t harvest RF energy at arbitrary distances (current prototypes: ≤ 10 m) • Diode drop limits energy harvesting HotPower ’08 — Ben Ransford — 22

CRFID applications • Medical implants [Oakland ’08] • RFID Sensor Networks [HotNets ’08] • Computation in inaccessible locations. fragile hazardous HotPower ’08 — Ben Ransford — 23

Future developments • Our work: • Fully implement checkpointing, reordering • Device profiling • CRFIDs: • Intel Research competition (Google intel wisp challenge ) HotPower ’08 — Ben Ransford — 24

Summary • Computational RFIDs: general-purpose batteryless computers • Mementos for forward progress • Checkpointing to cope with constant power interruptions • Program reordering to cope with fluctuating voltage HotPower ’08 — Ben Ransford — 25

Applications? Challenges? Alternatives? ransford@cs.umass.edu HotPower ’08 — Ben Ransford — 26